Step 1: Collecting Data:

Step 1(a):Saving the dataframe locally in a CSV format

Step 1(b): Fixing the same Tweets for the project by importing the CSV file that has been saved in step 1(a)

Step 2: Preprocessing the Tweets

Even after preprocessing the tweets, we can see that last word of every sentence does not make any sense. Removing the last word:

Removing stopwords

step 2(b): Labelling the tweets using Transformers Pre-Trained model

Setting neutral as 0, positive as 1, and negative as -1

Step 3: Feature Extraction

Step 3(a): Selecting k features from 12866 features

Step 3(b): Splitting data to train and test data

Step 4: Model Training

Step 4(a): Saving the model Locally for later use

Step 4(b): Loading the model that has been saved in step 4(a)

Step 5: Model Evaluation

Step 5(a): Predicting the X_test tweets sentiment for model evaluation

Step 5(b): Comparing y_test and y_pred to evaluate the model using different metrics